Lesson 60: Building Data Analysis Reports from CSV or Excel

A data analysis report is not “analyze this file.” It starts with a business question and ends with an actionable judgment.

Skills Used

Search skills.lc for data analysis, csv, visualization, or python. In this environment, use spreadsheets:Spreadsheets, Shell-based scripts, documents for reports, and presentations for executive summaries.

Workflow

Define the question first, such as “Why did conversion drop in the last three months?” The question determines metrics, segments, charts, and interpretation.

Build a data dictionary so every column has a meaning, unit, and business context. Then run exploratory analysis: trends, segments, anomalies, funnels, retention, and correlations.

Write the final report with background, data scope, key findings, evidence charts, recommended actions, risks, and next validation steps.

Human Review

Do not write correlation as causation. Do not hide filters. Do not ignore sample size. Do not choose only the charts that look good.

Exercise

Use an orders.csv file and ask OpenClaw to produce analysis.ipynb, charts, report.md, and an executive summary.

The value of analysis is not chart volume. It is clearer action.